Abstract
Tuberculosis (TB) is a lung disease that tops the world mortality rate. The widely employed method of diagnosis is the X-ray images which gives the pictorial information of the lungs. In the era of the Internet of Things (IoT), Artificial Intelligence (AI) using deep learning is among the most efficient methods in detecting lung-related diseases, and in classifying the related X-ray images. To trust the achieved decision, in this study ResNet-50 was used to classify TB and normal patients X-ray images, also, a Gradient-weighted Class Activation Mapping (Grad-CAM) was used to extract the features of the last pooling layer of the ResNet model to know what makes the model classify the X-ray images based on the given classes.